Similar-Tweeters

• A Django and ReactJS based web-application utilizing the Twitter API to find accounts with similar tweet styles.
• Modern NLP techniques such as neural language modeling, topic modeling, as well as general deep learning techniques.

Setup

Setup virtual environment.

In order for the program to run, we need to following to be set up properly.

pip install -r requirements.txt

Build Frontend from source

Build the static files from source. This allows it to loaded from the django backend. Use the followings commands:
cd text_backend/text_frontend

npm install

npm build

Migrate Database

This ensures that all of the data is where we want it. If the Django model is tweaked at all, this must be done before the results can be implemented. Use the following commands: cd text_backend

python manage.py migrate --run-syncdb

Setup Twitter API

There must exist a twitter_config.yaml file on the top level with the bearer token for the application. This is what allows the Twitter API to function. The YAML files should be of format such that the bearer token can be accessed via data[twitter][bearer_token].

Additionally, we want to have the top users pre-loaded into the database. We can use the following command:

python text_backend/preload_users.py

Train a Model

Clean the Data

In order to train a model, we must have data. If using the example book_data, this can be done via clean_book_data.py script. Otherwise it must be cleaned and put into .txt another way.

...